Damping Input for the the frequency response analysis 1

[vibrostics]

Some time back I was working on the Automotive Exhaust system frequency response analysis. I had been asked to use only the structural damping ratio rather than the viscous damping ratio. Can any one explain why? Is that to simplify the problem or the viscous damping effects are minimal?

Thanks.

San.

 

[GregLocock]

FEA or real world?

Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[vibrostics]

It is FEA.

Thanks.

 

[GregLocock]

and how well does FEA cope with viscous damping?

Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[vibract]

Well, literature tells that there are several kinds of damping models. Among them to carry out FEA calculatins, viscous damping has been chosen so that its energy dissipation per cycle of osciallation is equal to that of original damping present in the structure. Thus, for the ease of calculations the equvalent damping is chosen as Viscous damping represnting the whole damping of the system. Further, to uncouple the system of equations in FEM, this viscous damping is taken as the Rayleigh's damping model i.e C=aM+bK ( where first term represents the inertial damping matrix representing the energy loss associated with the change of momentum and second represents the stiffness damping matrix representing the loss of energy due to local deformation forces at joints near the mass elements - simplified form of linear structural damping).
Hence it's customory and a reason to use viscous damping model in FEM calculations assuming it has a Rayleigh's proportional damping model.
Regards

 

[vibract]

In continuation to the previous post..
To consider only the structural damping, it's requird to find other damping model values. i.e structural damping = overall damping value - (material damping values+ etc)
But, depending on your structure and type of application, type of energy dissipation, we can make some assumptions.
But generally, material damping contribution is very less compared to structural damping.

2nd case:
If the energy loss in your structure is mainly due to energy loss at joints, rubbing then we can neglect the first term in our Rayleigh damping model. i.e C=bK ( representing purely the energy loss at joints etc)

 

[vibrostics]

Thanks for the reply and explanation. I think that is the reason why the Frequency response analysis in Nastran uses the Structural Damping by neglecting the “aM” term in the Rayleigh's damping model. But it is also said that the Direct Transient response (time domain) analysis uses the Viscous damping input rather than structural damping. In Nastran it is internally done using the Structural damping value and the first global frequency. Could you explain why it is like that?

Thanks and Regards.

 

[vibract]

As damping is one of the natural charactarestics of your structure under analysis, the kind of domain matters nothing here. Whether it's time domain or frequency domain, damping remains same. As told before, because of the mathematical ease, viscous damping is considered as a representative factor for the whole damping. If we want to study the effects of only some special damping types, then its just the case of additions or substractions. ( further one thing we have to remember here, removal of one damping type doesn't mean that we completely removed it. This is because of the interdependency of all damping types within the overal damping. In mathematical sense, there is no linear relationship between the available damping types within the structure. Hence, it's very difficult to isolate one particular damping type compeltely)

 

[vibrostics]

Thanks for the Explanation. It was helpfull.